Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
NBER WORKING PAPER SERIES TARIFF LIBERALIZATION POLICY AND FINANCIAL RESTRICTIONS Joshua Aizeninan Working Paper No. 1253 NATIONAL BUREAU OF ECONOMIC RESEARCH 1050 Massachusetts Avenue Cambridge, M& 02138 December 1983 The research reported here is part of the NBER's research program in International Studies. Any opinions expressed are those of the author and not those of the National Bureau of Economic Research. NBER Working Paper #1253 December 1983 Tariff Liberalization Policy and Financial Restrictions ABSTRACT The purpose of this paper is to assess how restrictions on capital mobility affect adjustment to a tariff liberalization policy. This is done by comparing the adlustment process under free and restricted convertibility of foreign assets in a regime where the commercial exchange rate is pegged. It is shown that trade liberalization causes in the short run a larger drop in domestic goods prices and a smaller current account deficit in a regime with restricted convertibility. Similar results apply also for the long—run current account effects of the liberalization: they are smaller under financial restrictions. Joshua Aizenman 3718 Locust Walk, CR Economics Department University of Pennsylvania Philadelphia, PA 19104 215—898—7409 1. Introduction The experience of countries in the Southern—Cone has raised the question of the optimal way to implement liberalization policies. Countries there (and elsewhere) have used various restrictive policies on international trade in goods and assets. Commercial policies have resulted in distortions which have opened a wedge between the domestic and foreign prices of traded goods. Financial restrictions, on the other hand, have occasionally introduced a wedge between the exchange rates used for commercial and financial transactions. A country that wishes to liberalize its policies is therefore confronted with a choice among various liberalization schemes. A question under current debate is the desirability of imposing (or preserving) financial controls during attempts to liberalize commercial policies.1 A necessary step in answering such a question is understanding the effects of financial restrictions on the adjustment process. The purpose of this paper is to provide a framework capable of modeling those issues in a monetary economy. The paper will first focus on modeling two financial regimes. The first regime is a standard fixed exchange rate regime, in which a unified exchange rate is applied for financial and commercial transactions. In the second regime convertibility of domestic and foreign assets is effectively restricted by the policy maker. These restrictions can drive a flexible wedge between the exchange rates used for various transactions——a wedge whose magnitude is market determined. This might be the case under the formal use of multi—tier exchange rates1 or in an economy where there is significant use of a "black" market for foreign exchange. A further purpose of the analysis is to model —1-— such an economy. Next, the adjustment process to a commercial liberalizaiton policy is contrasted for the two regimes, allowing us to study the effect of financial restrictions on the process. The underlying framework is a version of the Calvo—Rodriguez currency substitution model, modified in several directions to allow for convertibility restrictions and for the inclusion of commercial policy. The analysis is conducted for the case where all prices that are not officially pegged are fully flexible. It can be readily extended to allow for slow wage adjustment. Section 2 presents the model underlying the analysis. Section 3 uses the model to evaluate the effects of financial restrictions on the adjustment process. Section 4 summarizes the findings, closing the paper with comments on the effects of introducing sluggish wage adjustment. This section also describes alternative uses of the model to analyze the effect of financial restrictions on adjustment to a devaluation and monetary policy. 2. The Model 2.a. The Case of Financial Constraints Consider a small open economy, subiect to commercial and financial policies. A minimal framework that is capable of describing it should at least allow for two goods and two assets. A simple example is the case where consumers allocate their wealth between domestic and foreign money balances, subject to convertibility restrictions. Such restrictions may introduce a wedge between the exchange rate that is relevant for commercial transactions and the exchange rate that applies for foreign assets. A useful way of modeling this outcome is the case of a two-tier exchange rate.2 Suppose that there is a given stock of foreign exchange in the hands of the private sector, —2— and the policy maker effectively controls financial capital flows. The restrictions on private capital flows imply that the exchange rate that clears the market for foreign assets can diverge from the commercial exchange rate. Suppose that the authorities peg the commercial rate at a preannounced rate, letting the market determine the financial rate. Let e and e denote the c f commercial and financial rates respectively (each defines the domestic currency equivalent of a unit of foreign currency). In such an economy the capital account is nil, and the balance of payments can be identified with the balance of trade. While such a formulation seems to be quite restrictive, it can describe effectively an economy where financial restrictions introduce a flexible wedge between the prices of foreign currency for various transac- tions. The magnitude of such a wedge is market determined, and will play an important role in our discussion.3 Financial liberalization can he modeled as either a policy that increases foreign assets in the hands of the private sector, or as a policy which abolishes the restrictions that caused the wedge between the various exchange rates. As in Calvo—Rodriguez (1977), the desired asset composition depends on the discrepancy between the returns on various assets.4 Equilibrium in the asset market can be summarized by: (1) where ef . M/N = 9(ef/ef), 0' > 0. and N denote private holding of foreign and domestic money balances. Domestic goods are taken to be an imperfect substitute for foreign goods. The small economy is facing a given price of foreign goods, denoted by * p • A tariff at a rate t results in a domestic price of foreign goods equal to e p*(1 + t). Let us denote by E the nominal expenditure, by S the share of expenditure devoted to domestic goods, and by X the foreign demand —3— for the domestic goods. The total demand for these goods is thus given by: * (2) E * (e(1+t) p /p) • —+ X(e • p /p) where 5', X' > 0, describing the dependence of demand on relative prices. Let Y denote the supply of home goods and R the tariff revenue, whose proceeds are assumed to be rebated in a lump sum manner.5 The domestic demand for imports is given by (l—)E/{(l+t) • ec •*]• Thus, the tariff revenue is given by (3) R = t • * e • p ( (l—)E * +t)e.p C = _________ 1+ Expenditure is a function of nominal disposable income (PY + R) and wealth (V) (4) E = E(PY + R, V) where 0 < E1, < 1, E2 > 0, and nominal wealth is given by (5) V = M + ef • N All prices that are not officially pegged are assumed to be fully flexible. This allows one to proceed by assuming that output is at its full employment level, denoted by Y0, and goods prices are set such as to clear the goods market: (6) Y0 =.-+X p —4— We assume the absence of an active monetary policy. Thus, money balances accumulate over time via the balance of payment surplus. In our two—tier regime such a balance of payment surplus is equal also to the current account.: (7) M = p X — • (1—)E l+t Using the goods market equilibrium condition (eq. 6) we find that = p (8) Y • + R — E(p • Y + R,V). For a given level of foreign assets (M*) eq. 1 and 8 provide the motion rules for our system while eq. 6 defines the goods prices for given values of (ef M). Thus, long run equilibrium values of wealth (V) and goods prices (p) are obtained from: (9) p • Y (10) y = + R = E(p 0 . 0 . Y 0 + R, V) + x. p For the equilibrium values of V and p the long run equilibrium financial exchange rate is given by: (11) ef = * 8(0) • V M (1 + 0(0)) From eq. 1, 6, 8 we obtain that our dynamic system can be described by —5— f(M, (12) ef = (13) M = h(M, ef) > 0 fi < ef) h1 < 0, h2 < 0. As is evident from eq. 12—13, a linearized version of our system exhibits saddle path stability around the long run equilibrium, and the dynamics of adlustment are described in Figure 1, where AA corresponds to the saddle path. 8 Curve = 0 describes combinations of money balances and the financial exchange rate that are compatible with anticipation of a stable financial exchange rate. An increase in money balances will raise the demand for foreign assets. To regain composition equilibrium at = 0, the financial rate should depreciate at the same rate that the money balances rise. Thus, ef = 0 is upward—sloping, with a unitary elasticity. In a similar way, curve N = 0 describes combinations of (M, ef) that are consistant with current account equilibrium. Inspection of our system reveals that around the equilibrium N = 0 is consistent with an adjustment of (M, ef) that leaves wealth intact. Thus, = 0 is downward—slooing. 2.b The Case of Free Convertibility Under free convertibility there is a unified exchange rate. Assuming a fixed rate regime, the authorities use reserves to peg the exchange rate for all transactions. This implies that any portfolio disequilibrium is corrected by means of a swap of domestic money balances with foreign assets. Such a swap does not alter wealth (V), it only affects the private sector's portfolio —6— composition.9 Thus, we cannot identify the current account with the change in money balances. Instead, the current account is equal to the net accumulation replaces : of wealth. Thus, in eq. 7 and 8 = p (8') • Y + R — E(p • Y + R,V) As before, eq. 9—10 define the long run values of wealth and goods prices. Eq. 11, however, now defines the equilibrium holdings of foreign assets for the given exchange rate (ef = e). A useful property of our model is that eq. 10—11 hold for both regimes, defining the long run equilibrium. From eq. 6, 8' we obtain that the adjustment process under a fixed rate regime is given by (14) 7 = g(V) , g' < 0. From eq. 1 we also obtain: V = (1 + (15) Notice 9(0)) • that for a given wealth eq. 6 gives us domestic goods prices. 3. Adjustment to a commercial liberalization policy We would like to use our model to contrast the adjustment to a commercial liberalization policy under the two regimes. This will enable us to assess the economic relevance of financial restrictions, because under a two—tier regime we impose such restrictions, whereas under a unified exchange rate they —7—. are nil. To make the comparison meaningful, consider the case of an identical initial long run equilibrium under both regimes. Suppose that by the choice of units in such an equilibrium p = ec = ef = • 10 For simplicity of exposition we consider now the case of a low initial tariff, such that in the new liberalized regime t = 0. Because eq. 9, 10 define the long run equilibrium for both regimes, it is useful to start by deriving the long run effects of the commercial liberalization: (16) (17) — dt (1—) + > 0 L.R. S' dV = (1 - E1)Y L.R. [(l-)+ ' + l-] > . 2 0 For given domestic goods prices the liberalization switches demand from domestic to foreign goods. To regain equilibrium domestic goods prices should drop. The resultant drop in the price level induces also a drop in the demand for nominal wealth resulting in a lower equilibrium V. While the drop in nominal wealth is the same under both regimes, its decomposition between quantity and price adjustment differs. From eq. 11 we obtain that in a two— tier regime the long run effect of the liberalization on the financial exchange rate given by: de —* L.R. N (1 + 0(0)) dt Thus, the liberalization will generate a drop in the equilibrium —8— (long run) financial exchange rate. Furthermore, from eq. 11 we find that the cumulative current account adjustment under the two regimes is given by under a fixed exchange rate regime, whereas under a two—tier regime L.R. dV 1 • it equal to Thus, we can concludes that, in the long 1 + 010) d L.R. run, financial restrictions are manifested in shifting part of the adjustment is from quantities to prices, implying a smaller current account adjustment. The impact effect of the liberalization can he read from eq. 6. Under both regimes we obtain that = (19) S.R. '+ (1 — cS) + E2 S+6' +—y— 5E Under a fixed exchange rate regime the liberalization policy does not = have direct wealth effects in the short run (thus 0). S .R. This is not the case, however, under a two—tier regime. When the liberalization goes into effect, the new long run financial rate is expected to drop. This anticipation will reduce the demand for foreign assets, resulting in a drop in the financial exchange rate. The adjustment process can he traced with the help of Figure 1. Let 'a' denote the initial long run equilibrium that was common to both regimes. Notice that the liberalization policy does not effect the locus of ef = 0. A lower tariff implies a lower price level. This will increase real wealth, inducing at a given financial rate a current account deficit. Thus, if we wish to preserve a current account balance (tt = 0 for a given M) the financial exchange rate should appreciate enough to eliminate the current account pressure. Thus, following the tariff reduction curve M = 0 shifts downward. Because the liberalization —9— policy does not affect the locus of ef = 0, the saddle path shifts downward. The new long run equilibrium under a two—tier regime is obtained at point c. The effect of the liberalization is to place us under perfect foresight on the new saddle path (point b), below the initial equilibrium. Thus, in the short run the financial rate will appreciate, undershooting its long—run value. The effect of the liberalization is also to generate a current account deficit. During the adjustment to the new long run equilibrium money balances will drop, the financial exchange rate will appreciate, and goods prices will further decline to preserve the market equilibrium of assets and goods. In terms of Figure 1, we will move gradually along the saddle path from b to c. As eq. 19 reveals, goods prices will drop in both regimes when the liberalization goes into effect. Under financial restrictions the appreciation of the financial rate following the liberalization will magnify dV > the drop in goods prices (because S.R. T.T. dV = 0). S.R. FIX Under a fixed exchange rate regime the long run equilibrium is obtained at point c'. Because the exchange rate is unified and fixed, we will stay in the short run at point a, and gradually will move towards the new long run equilibrium. In both regimes we observe a current account deficit following the liberalization. This deficit can be shown to be smaller under a two—tier regime. That is because the instantaneous appreciation of the financial rate will substitute partially for the needed quantity adjustment achieved via the current account. The resultant wealth adjustment under a two—tier regime partly cushions the current account deficit. Again, we observe a trade—off —10— between quantity and price adjustment. Notice that our analysis shows that current account deficit will accompany an appreciation of the financial exchange rate. This is in contrast with the analysis of the asset approach for floating rates, where it is shown that a current account deficit is accompanied by a depreciation of the financial exchange rate. [See Dornbusch and Fischer]. The different correlations among the exchange rate and the current account reflect the underlying differences between the two exchange rate regimes. Under a floating rate system, a current account deficit reduces holdings of foreign assets by the private sector, leaving holdings of money balances intact. Under a two—tier system a current account deficit reduces holding of money balances, leaving holdings of foreign assets in the hands of the private sector intact. To regain asset composition equilibrium we need depreciation in the first case, appreciation in the second. In terms of Figure 1, under floating * rates foreign assets (M ) would replace domestic money balances, and the curve describing 0 would be downward sloping. It can be shown that although the slope of M = 0 might be either positive or negative, the saddle path under floating rates slopes downward. 4. Summary and Concluding Remarks The foregoing analysis has assessed the effect of capital mobility restrictions on the adjustment process to a tariff liberalization policy. This was done by comparing the adjustment process under free and restricted convertibility of foreign assets in a regime where the commercial exchange rate was pegged. In a regime with restricted convertibility we observe a larger drop in —11— domestic goods prices and a smaller current account deficit when the trade liberalization becomes effective. This is accompanied by appreciation of the financial rate. During the transition to the new long run equilibrium we observe in both regimes a further drop in goods J)rices and C4. current account deficit. Financial restrictions reduce the long—run effects of trade liberalization on the current account. This analysis can he readily extended in several (jirections. First we can extend our framework to allow for short—run wage rigidities. This implies that the effect of the liberalization Is to generate unemployment in the short run. The magnitude of the resultant unemployment will he larger under financial restrictions, because of the larger drop in goods prices in the short run. Nominal wage rigidities imply also that a "smart' nominal policy, such as devaluation, could reduce the resultant unemployment. If wages are rigid in the short run, financial restrictions enhance the importance of such measures. 11 The paper's framework also permits an assessment of how financial restrictions affect adlustment to devaluation, a monetary injection, and other policies. It can he shown that the effect of financial restrictions is to increase the current account's short—run response to devaluation and monetary 12 pol icy. — I- FOOTNOTES 1. For a survey of the literature on the timing and order of liberalization policies see Edwards. Earlier work on those issues can be found in McKinnon. 2. For models of two—tier exchange rate regimes see, for example, Flood and Marion, Marion, Bhandari and Decaluwe and the references thereupon. 3. The model can be also applied for the case where part of the financial transaction uses the official rate, and another part uses the "black market" exchange rate. Our analysis neglect issues related to interest payments, because the service account seem to play insignificant role in explaining the short run adjustment process to a liberalization policy. The model can be modified to include world bonds instead of foreign exchange (see, for example, Dornbusch and Fischer). 4. Similar models were used intensively in the formulation of the asset approach. For example Dornbusch and Fischer analyze exchange rates and current account, whereas Eichengreen applies a related model to analyzing tariffs under flexible exchange rates. 5. The model abstracts from issues related to fiscal policy. 6. Notice that E denotes nominal expenditure. Making real expenditure depend on real disposable income and real wealth would not affect the results reoorted in the paper. 7. The dynamics of our model can be viewed as consistent with the permanent income or life cycle hypothesis, according to which assets are used to smooth consumption over time. 8. The positive slope of the saddle path is the result of the signs of the —13— partial derivatives of eq. 12—13. 9. We assume the absence of an active monetary policy, and that the private sector ignores the composition of the central bank's balance sheet in arriving at portfolio decisions. For a discussion about assets swap under a fixed rate regime see Frenkel and Rodriguez. 10. Our analysis can be readily extended for the case where at the initial equilibrium the financial rate exceed the commercial rate. 11. For example, consider the case where employment (L) is demand determined, L = L(w/p) The wage is pre—set1 where w is the money wage, L' < 0. adjusting according to a rule = — L)/L , where L is the full employment level. In such a framework the total unemployment during the adjustment process is proportional to the unexpected domestic goods price drop at the liberalization day. 12. Notice that the short—run effect of financial restrictions was to reduce the current account deficit following tariff liberalization (in comparison to the current account under a unified fixed rate). This is because a tariff policy has real long run effects, whereas a monetary policy and devaluation prove to be neutral in the long run in our framework. —14— Re ferences Bhandari, S.J. and B. Decaluwe, 1983, Stochastic Implications of Incomplete Separation Between Commercial and Financial Exchange Markets. Unpublished Manuscript. Calvo, G. C.A. Rodriguez, 1977, A Model of Exchange Rate Determination Under Currency Substitution and Rational Expectations, Journal of Political Economy 85, 617—625. Dornbusch, R.and S. Fischer, 1980. Exchange Rates and the Current Account, American Economic Review 71, 960—971. Edwards S., 1983, The Order of Liberalization and the Current and Capital Accounts of the Balance of Payments: A Survey of Malor Issues. 'Working Paner, Dept. of Economics, UCLA. Eichengreen B.J. 1981, A Dynamic Model of Tariffs, Output and Employment Under Flexible Exchange Rates. Journal of International Economics 11, 341—360. Flood, R.P. and N.P. Marion, 1982, The Transmission of Disturbances Under Alternative Exchange Rate Regime With Optimal Indexing, Quarterly Journal of Economics 95, 45—62. Frenkel, J. and Rodriguez, C., 1975, Portfolio Equilibrium and the Balance of Payments: A Monetary Approach, American Economic Review Vol. 65, December, pp. 674—88. Marion, N.P. 1981, Insulation Properties of a Two—Tier Exchange Market in a Portfolio Balance Model, Economica, 48(189), pp. 61—70. McKinnon, R.J., 1973, Money and Capital in Economic Development, Brookings Institution. —15— e <1 ef= 0 M =0 C' A- M FIGURE 1